PROJECT SUMMARY Current therapy for acute myeloid leukemia (AML) is inadequate and the overall survival rate for patients with the disease is still poor. Despite initial response to standard therapy, many patients eventually relapse from their diseases. The leukemia stem cell (LSC) model proposes that LSCs are at the apex of the AML patient samples and are capable of initiating and maintaining the disease. LSCs are rare and largely more quiescent than the bulk leukemia, properties speculated to underlie drug resistance and disease relapse. Eradicating LSCs would therefore be important to ensuring long-term remission. In a previous study, we showed that JMJD1C, a member of the lysine demethylase 3 (KDM3) family, is one of the top differentially expressed MLL- AF9 target genes in the mouse MLL-AF9 LSC versus its normal counterpart, as well as in human MLL- rearranged (MLLr) versus non-MLLr leukemia. Using a conditional knockout model, we demonstrated JMJD1C is required for AML LSCs in both murine models of MLL-AF9 and Hoxa9/Meis1 leukemias, but dispensable for normal HSC function, therefore a potential therapeutic target for eliminating LSCs. In this proposal we plan to examine the overall hypothesis that JMJD1C is required for human AML LSC function through its demethylase activity and can be therefore targeted together with other AML therapies to achieve better therapeutic outcome. In a preliminary study, we used a CRISPR/Cas9 mediated negative selection screen and identified the catalytic jumonji domain (JmjC) and zinc finger domain (ZFD) as important for leukemia cell survival and showed that they are important for MLL-AF9 leukemia in vitro and in vivo. To assess the therapeutic potential of JMJD1C in human AML, we will i) use the CRISPR/Cas9 to mutate JmjC and ZFD respectively in primary patient samples and examine the role of JMJD1C in human LSC function both in vitro and in vivo using xenotransplant models; ii) examine whether combining standard therapy and JMJD1C mutation would improve therapeutic efficacy in pre-clinical xenotransplant models. To understand the mechanism of JMJD1C function in human LSCs, we will i) examine the enzymatic activity of JMJD1C. The question still remains as to the precise enzymatic activity of JMJD1C. Based on our preliminary data, we plan to use a combination of biochemical and epigenomic approaches to determine whether JMJD1C is a demethylase for H3 lysine 36 methylation in vitro and in vivo. In addition, we will investigate the role of ZFD in JMJD1C's enzymatic activity. ii) our preliminary study identified the classic repressor complex NCoR complex as JMJD1C interacting proteins. We plan to investigate whether JMJD1C's enzymatic activity as well as function in AML LSC is mediated through association with NCoR complex. Overall, our proposal once accomplished will establish a role and yield mechanistic insights of JMJD1C in human AML LSC function, and provide validated target domains that can be used in the design of novel therapeutic interventions in AML.